Method and system for semantic affinity search

Abstract

A search interface and searching technology for interactive search refinement based on feature similarity (ISRFS), which provides an interactive search tool based on human associative and semantic knowledge. Given a query, the tool may retrieve the most suitable textual items in a repository of textual data that is categorized or / and partially tagged. In the user interface, an equalizer controller may be used to refine the search results, for example, based on manually or automatically extracted features and a graded semantic scale to adjust the weights of the various features. The technology may be applied to any data repository.

Description

CROSS-REFERENCE TO RELATED APPLICATION[0001]The present application claims benefit of U.S. Provisional Patent Application Ser. No. 60 / 996,990, filed Dec. 13, 2007 and entitled Method and System for Semantic Affinity Search, the entire contents of which are incorporated herein by reference.BACKGROUND OF THE INVENTION[0002]In the current state of internet search technology, the keywords typically must appear in the title or description of the item being searched in order to be located and listed as a search result. Moreover, in such a search, extra or additional keywords (search terms) entered in order to better define the searched items often impair or unnecessarily constrict the results, sometimes even returning no results at all (zero matching) due to the Boolean nature of the search.SUMMARY OF THE INVENTION[0003]Some embodiments of the present invention may introduce a textual search technology, which may be called “interactive search refinement based on features similarity” (ISRF...

AI Technical Summary

Benefits of technology

[0003]Some embodiments of the present invention may introduce a textual search technology, which may be called “interactive search refinement based on features similarity” (ISRFS). ISRFS may be an interactive refinement search tool that may be based on human associative and semantic knowledge, which may address, for example, two aspects of searching. One aspect may be the ability of a search engine to automatically index large collections of textual items according to their main semantic concepts. Another aspect may be the ability to provide a good estimator that may measure semantic affinity between any textual item and a collection of predefined semantic target features. A system, according to embodiments of the present invention, may meet, for example, both of these requirements and in some embodiments together with a graphical user interface (GUI) that may allow the user to easily control the affinity quality and the search process. Given a query, a tool according to embodiments of the invention may retrieve the textual items having high affinity to the user's searched item from a repository of textual data. An equalizer controller may be used to refine the search results based on, for example, semantic similarity, automatically extracted features, and a graded semantic scale, known as semantic affinity metric (SAM).

[0005]According to embodiments of the present invention, the TC technology may be modified to provide the user with an enhanced search experience and a user-friendly interface that may enable the user to easily define target features and their respective weights, which may be gradually changed to have any value between the lower and upper limits of a range of weights. According to embodiments of the present invention, the user may easily control the search outcome with higher semantic resolution and graded semantic scale as compared to what the Boolean nature of the TC technology may otherwise permit. According to embodiments of the present invention, ISRFS technology may provide a graded matching level between the target of the search and the textual items being searched through. A user may define the item s / he is looking for to have “more of feature A” and / or “less of feature B,” etc.

[0006]A method according to aspects of the invention may take advantage of both the core algorithmic level and the unique user interface. A user or a consumer search tool application may be used throughout some of the following explanations merely as an example to illustrate the technology, which, in general, may be applied to any textual repository. The semantic affinity search may consist of two steps. A first step may be a semantic search and a second step may be a semantic refinement. This two-step procedure may reflect an intuitive and efficient search strategy.

Problems solved by technology

Moreover, in such a search, extra or additional keywords (search terms) entered in order to better define the searched items often impair or unnecessarily constrict the results, sometimes even returning no results at all (zero matching) due to the Boolean nature of the search.

Images